oracle-dba
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and analyze data from database performance views (e.g., v$sql, v$session) and provides the agent with powerful capabilities to modify infrastructure via OCI CLI (e.g., scaling, deleting databases). This creates a surface where malicious content embedded in database metadata (such as SQL comments or session info) could potentially influence the agent's behavior.\n
- Ingestion points: Performance and metadata views including v$sql, v$session, and v$system_event (described in SKILL.md and references/sqlcl-workflows.md).\n
- Boundary markers: None present in the instructions to distinguish between data and instructions.\n
- Capability inventory: Extensive administrative capabilities including OCI CLI commands for database creation, deletion, scaling, and lifecycle management, as well as SQL execution via SQLcl (references/oci-cli-adb.md, references/sqlcl-workflows.md).\n
- Sanitization: The skill does not define explicit sanitization or validation steps for the metadata retrieved from the database before it is used to inform further actions.
Audit Metadata